1. Field of the Invention
The present invention relates to an automatic fault diagnostic network system and an automatic fault diagnostic method for networks and, more particularly, to an automatic fault diagnostic network system and automatic fault diagnostic method for networks, which automatically diagnose a fault in a network element as a management target at the time of occurrence of the fault.
2. Description of the Prior Art
In general, a network management system for managing a network is connected to it, and the overall structure of the network as a management target and the connection states of network elements are displayed on a built-in display means. A network manager can check through this display means whether, for example, the network is normal or any fault has occurred in any network element.
In recent years, networks have rapidly increased in size and spread over wide areas. When a fault occurs in a network element, it is important to immediately take necessary measures to eliminate this. For this reason, demands have arisen for means by which network managers can properly cope with faults on networks. As such a means, a network management system is constructed from a fault log creation/storage means for automatically detecting a network element in which a fault has occurred, automatically creating a fault log on which at least the name of the network element and the type of fault are recorded, and recording the log, a trouble ticket issuing means for issuing a trouble ticket on which the detailed information recorded on the fault log can be recorded, and a trouble ticket storage means for storing the trouble ticket on which the detailed information is recorded. With such a system, an attempt has been made to improve network management performance at the time of occurrence of a network fault.
The anti-network fault management system disclosed in Japanese Unexamined Patent Publication No. 06-326751 is an example of this system. FIG. 1 is a block diagram showing the function/arrangement of the network management system disclosed in this reference. The arrangement and function of a conventional network management system will be described with reference to FIG. 1. A general network management system includes a display unit 101 for displaying a network as a management target in a diagrammatic form, a control section 102 which is connected to the display unit 101 to control the contents displayed on the display unit 101 or controls an operation unit 104 and database 103, the operation unit 104 which is connected to the control section 102 to provide it with a signal for designating display contents, and the database 103 which is connected to the control section 102 to store the contents of the network which are to be displayed.
Such a network management system is connected to the network constituted by network elements as management targets through a connection unit. When a fault occurs in one of the network elements, this system displays fault information on the display unit 101, creates a fault log, and stores it in the database 103. In the above reference, in particular, there is disclosed an idea that enables a user to quickly find a solution method for a new fault at the time of occurrence of the fault upon referring to a fault log recorded in the past by issuing a trouble ticket on which measures taken against faults by the network manager are also recorded.
According to conventional techniques associated with network management systems, as in the above example, the network management performance can be improved to some extent by issuing or storing a trouble ticket on which a fault log is recorded/updated. With the conventional techniques and methods, however, an improvement in network fault diagnostic function is undesirably limited, and basic problems in network management remain unsolved. The first problem in network management by the conventional network management system is that a network manager must acquire sufficient experience associated with diagnosis because the system has no diagnostic function of acquiring appropriate fault information that enables a search for the cause of the fault, although the system allows the manager to check the occurrence of the fault and its contents to some extent. This problem arises because there is no means that automatically diagnoses a network element in which a fault has occurred.
The second problem in network management is that the network manager must perform determination or operation by himself/herself to check whether a network element in which a fault has occurred can be diagnosed. If, therefore, it is unclear whether each network element has a diagnostic function, or a network maintenance person without any know-how about diagnosis is to perform operation, this problem brings great difficulties in taking countermeasures against network faults.
The third problem in network management is that even if a means for executing diagnosis is prepared because of the necessity of the execution of diagnosis for a search for the cause of a fault, a maintenance person must check a fault in a network element first, and then give the network element an instruction to execute diagnosis from a system other than this network management system. As described above, the conventional network management system has no effective means for taking proper measures against a fault on a network and reducing the work load on a maintenance person.
It is therefore required to automatically find the cause of a fault on the basis of the fault and diagnosis.
It is an object of the present invention to provide an automatic fault diagnostic network system and an automatic fault diagnostic method for networks, which can automatically find the cause of a fault on the basis of fault diagnosis.
It is another object of the present invention to provide an automatic fault diagnostic network system and automatic fault diagnostic method for networks, which can automatically find the cause of a fault on the basis of fault diagnosis by performing autonomous determination on a detected fault instead of making a network manager perform determination and operation to check whether diagnosis of a network element in which the fault has occurred can be made.
In order to achieve the above objects, according to the present invention, there is provided an automatic fault diagnostic network system comprising an automatic diagnostic system and a connection unit for connecting a plurality of network elements to the automatic diagnostic system, wherein the automatic diagnostic system comprises a fault detection unit for detecting an individual fault, and a test unit for running a test to find a cause of the individual fault in the network element in which the individual fault has occurred. This system actively searches for the cause of an individual fault by a test as well as detecting the fault, thereby automating fault processing.
In addition, the system includes a database, and the database stores a plurality of fault data corresponding to a plurality of faults. This allows the test unit to precisely test and determine the cause of the individual fault on the basis of the fault data corresponding to the individual fault detected by the fault detection unit.
The fault data includes the type of network element and fault contents. In many cases, the types of faults that occur in elements are known in advance. The contents of such faults are registered in the database in advance. The contents of a newly detected fault are registered in the database in correspondence with an element every time it is detected.
Information that can specify the contents of a fault is a correspondence between the type of element, e.g., the model or name of the element, and the fault contents. Such a correspondence table is formed in the database. Identical faults occur in elements of the same type. Certain faults occur in only specific types of devices. It is preferable that fault data further include diagnosis execution enable/disable information and diagnosis execution necessity information. With regard to a fault for which no diagnostic test, i.e., test, can be performed, information indicating that diagnosis cannot be executed is preferably prepared. It is also important to have diagnosis necessity information because some fault need not be diagnosed even though it can be diagnosed.
An automatic fault diagnostic method for networks according to the present invention comprises the steps of detecting a fault in a network element, making the fault correspond to fault contents in accordance with the type of element, and confirming the fault contents by testing the fault in accordance with the fault contents. Confirming fault contents is equivalent to finding the cause of a fault. If the cause of a fault is detected, the fault can be recovered by using a conventional means.
As is obvious from the above aspects, the automatic fault diagnostic network system and automatic fault diagnostic method for networks according to the present invention can specify the cause of a fault by executing a test for diagnosis of a network element in which the fault has occurred. If the cause of the fault is specified, the element can be recovered. It can also be determined whether diagnosis can be performed. In addition, even if diagnosis can be performed, it is determined whether diagnosis is necessary. Thereafter, a diagnostic test is executed.
The following effects can be additionally obtained.
(1) When a fault occurs in a network element, processing up to diagnosis can be automatically executed without any manual operation.
(2) Both the fault information and a diagnosis result can be quickly checked. This reduces the amount of work performed by a maintenance person.
(3) Since a diagnosis result can be analyzed by extracting data from the database, appropriate and sophisticated measures can be taken against a fault.
(4) With the above effects, the processing speed and accuracy in network management can be increased regardless of the state of a network and the skill of a maintenance person.